The fact that the Sun goes through cycles of activity was first noticed in the mid-19th century when astronomers began charting the number of cool, dark patches or "sunspots" on the solar surface where magnetic activity is intense. However, in the mid-1970s -- when the first accurate data of the Sun's polar magnetic field started being recorded -- astronomers found evidence for a possible link between the value of the field at the end of one cycle and the overall strength of the next cycle. If such a link does exist, one could then predict the strength of future cycles based on past data. Indeed, the polar field is so low in the current solar cycle that the next cycle is predicted to be the weakest for 100 years.

Rather than relying on observational data to predict the strength of future cycles, however, two US physicists last year tried to predict the activity of solar cycles by theoretically modelling the Sun's internal "dynamo action", which converts the turbulent movement of solar material into magnetic energy. Strangely, the model, which was developed by Mausumi Dikpati and Peter Gilman from the National Center for Atmospheric Research in the US, predicted that the peak activity of the next cycle would be very high -- quite at odds with the method using observational trends.

Arnab Rai Choudhuri and colleagues from the Indian Institute of Science in Bangalore and the Chinese Academy of Sciences in Beijing now think that Dikpati and Gilman's model was flawed because of the way it treated a key element of the dynamo model known as the "Babcock-Leighton mechanism", which produces a weak "poloidal" field when pairs of sunspots decay. Dikpati and Gilman fed records of past sunspot data into their model to calculate the strength of this mechanism, but Choudhuri's team say that the production of the poloidal field is inherently random and cannot be calculated in this way. Instead, they have created a model that is "corrected" around the minimum by polar magnetic field data, which they say is more accurate.

Using this model, Choudhuri's team calculate that the next cycle (known as cycle 24) will be about 35% weaker than cycle 23. If they are correct, this would be good news for satellite communications, because solar flares and coronal mass ejections, which can generate violent magnetic storms in the Earth's atmosphere, become more common as solar activity increases. But Choudhuri is happy to wait to see if he is right. "We shall have a verdict from the Sun-god himself within the next four to five years whether [Dikpati and Gilman's] prediction or our prediction comes closer to the truth," he said in a recent review paper.